Post Go back to editing

LT5560 oscillator

Category: Hardware
Product Number: LT5560

Hello, we would like to replace discontinued product SA602AD (from NXP). I checked your website, compare your products and I think that best choice is the mixer LT5560, because it has lower quiescent current that other. I would like to ask you if you have any recommendation for an oscillator in fixed frequency with low power? 

It will be use in fixed 45MHz RF input frequency of mixer,  output frequency of mixer will be 10.7MHz (fixed too). So I need LO input 55.7MHz or 34.3MHz oscillator. Do you have any recommendation for connection with 3rd overtone crystal or something similar? SA602 has this oscillator circuit:

 


I could use ADF4360-9 (I use in other project) but the ADF4360-9 have high quiescent current and I need something lower (And ADF4360-9 has frequency band 65-400MHz). Is it possible use for example TCXO crystal oscillator with clipped sinus wave output? 

Thanks, best regards. Tomas. 

  • Hi ,

    LT5560 should work for your application, but check the datasheet published gain, IIP3, and Noise Figure specs, especially page 21. I think you'll find the conversion gain of LT5560 to be somewhat less than SA602, but if the next stage can be higher input impedance at the IF=10.7MHz, i.e. >> 50 Ohms, then that higher load Z should result in higher conversion voltage gain. 

    LT5560 doesn't have the internal gain stage to support either fundamental or overtone xtal osc, but the internet is full of simple 1-transistor Colpitts oscillator circuits that should work well. We would generally recommend the high-side inject, for more favorable spurious performance, provided the system performance can accommodate the resulting spectrum flip. LT5560 will work fine with single-ended LO drive from an external crystal osc.    -Bruce H.

  • Hello, I would like to ask you a one question. I would like to use the LT5560 as a downconvertor a next stage will be monolitic crystal filter (freq 10.7MHz, BW +-7.5kHz, impedance 3000 Ohm). Is it better solution matched the output stage of LT5560 to 50Ohm and matched crystal filter input impedance from 3000 Ohm to 50 Ohm. Or is it better solution matched output of LT5560 to 3000 Ohm? Thanks Tomas. 

  • Hi Tomas,

    See datasheet figures 9 and 10, showing mixer output equiv. ckt. as already up around 1.2k ohms, with some shunt capacitance. Matching mixer output down to 50 ohms is convenient for test purposes, but then filter must be matched back up to 3k. Having 50 ohms interconnects helps simplify the testing with off-the-shelf 50 Ohm test equipment, but all the down and up impedance translation is going to be lossy, and probably add to the component cost. If you are putting mixer and IF xtal filter on one PCB, consider keeping the 10.7MHz mixer to IF filter connections at high impedance, which will require very short interconnections for low capacitance, and also keep it well shielded. If the filter is single-ended input, you will need some type of balun to maintain best overall performance. The mixer output shunt capacitance can be resonated out with some parallel inductance, or the balun itself can be resonant and also resonate out the mixer output capacitance, for lowest overall parts count. There's a lot of possibilities here, depending on the desired level of performance, and how much design time one wishes to put into the effort, and how important it is to reduce cost. Unfortunately I don't know of any 'canned' designs that illustrate this application, and we're not able to provide any deeper level of design assistance. 

  • Hi Bruce, thank you for your recommendation and tips. In past I designed the mixer with LT5512 (RF input 73-84MHz and IF 45MHz) and I choiced matched the output of LT5512 to 50Ohm and matched IF 45MHz crytal filter to 50Ohm (filtr has 1000 Ohm input). And there was about 3.5dB loses, but there was easily for tuning and design as you wrote.